Competitive SNSF MAPS Fund Awarded to ARTORG's Medical Image Analysis Laboratory Project "A-BEACON"

The Swiss National Science Foundation (SNSF) opened a call for proposals, open to all research areas, for the Multilateral Academic Projects (MAPS) fund on 01 April 2024. MAPS was created as part of the Second Swiss contribution, a mandate given by the Swiss Agency for Development and Cooperation (SDC).

Up until its closing on 01 July 2024, 334 proposals were received of which 29 were approved after review, the success rate coming in at almost 9%. Among the winning proposals were four from the University of Bern, including ARTORG's own Prof. Dr. Mauricio Reyes, head of the Medical Image Analysis Laboratory. We spoke to him about the project, what it means to have received this fund, and what he hopes to achieve in the coming four years.

Successful Funding for the A-BEACON Project at ARTORG

"A-BEACON" is an ambitious project tackling the unmet clinical need for more efficient, accurate, and reliable methods to assist clinicians in diagnosing and monitoring brain tumors. A-BEACON - which stands for AI-based Brain MEtastases TrACking and SegmentatiON - is a system designed to fill that gap, improve on existing AI-based models, and ultimately significantly reduce the time-consumption and labor-intensity of accurately analyzing brain MRIs.

The successful funding of this project came as a welcome surprise to Mauricio Reyes. "I wasn't expecting this grant to be accepted as I was told it would be very competitive," he admits. Reyes' project was selected together with three others from the University of Bern and 25 more projects from across the Swiss academic landscape. To be among the 9% of recipients is no small achievement for the researcher: "I think it's an honor. I think it's also a reflection of that emphasis we have at ARTORG, of performing research while targeting potential future translation, to see technologies beyond the demo. And I think that is, for us, a very nice surprise, that the project was perceived positively." The project was awarded 1'019'505 Swiss francs for the span of four years, with the conclusion of the project scheduled for the end of May 2029.

A Collaboration across Countries, Cultures and Clinics

A-BEACON is not a project exclusive to ARTORG and Mauricio Reyes' team. MAPS is a multilateral initiative which supports collaborative research projects between researchers in Switzerland and their colleagues in Bulgaria, Croatia, Hungary, Poland, and Romania. Reyes is joined in his work on the project by Prof. Dr. med. Claudiu Matei (Department of Neurosurgery, University of Sibiu, Romania), Assist. Prof. Dr. med. Jacek Kunicki (Maria Sklodowska Curie National Research Institute of Oncology, Warsaw, Poland), and Prof. Dr. med. Elitsa Encheva (Radiotherapy Department, St. Marina University Hospital, Medical University Varna, Bulgaria).

Although the reserachers had not previously met, there is great potential in the cross-cultural nature of the project. "I didn't know them before. I have seen their names here and there, but this is the first time we are engaging in such a project," explains Reyes. "The moment I presented it, they really loved it and were very excited to jump into the project. What makes this special is that AI technologies need to be evaluated in very different setups. And that's the case here; you have different countries with different resources. And you want those technologies to, ultimately, reach the patient through those doctors using the technology."

Indeed, the intercultural collaboration provides an ideal environment for the development of such technology. Although European guidelines regarding treatment processes exist, Reyes knows from experience that every medical center or facility develops their own workflows and protocols. This works in great favor for the project, as varying setups with nuanced implementation of protocols and guidelines provide an ideal testing ground for the to be developed AI model. While the primary development of the AI will be conducted through ARTORG at the University of Bern, the evaluation of the tool is planned as a multi-center evaluation across the different collaborating European countries.

"I think for us it's a great opportunity because typically you would have an evaluation that is, let's say, in Bern, or maybe a colleague in Zurich might say, yes, let's test it. But then it's a similar setup," elaborates Reyes on the evaluation process, "but when you talk about a different hospital, and that's something we will discover in the other countries, you have different, not only resources, but also mindsets, and maybe clinicians are taught differently. So, I think that's a great opportunity."

A Hope for the Project and Future of AI in Medicine

The received MAPS grant can fund the project for a period of four years. Much is planned for that timeframe and Prof. Reyes has high hopes for the development and realistic implementation of A-BEACON in clinics in the future.

One of A-BEACON's ambitions is to succeed where previous solutions to the time-consumption of manual segmentation of brain metastases from MRI scans have failed, and to overcome the significant hurdles existing AI-based models of similar utility are facing. What his team aims to achieve with A-BEACON is a system with Zero-Miss Detection, so 0% oversight, and a very small percentage of over-detection: "Picture a system that catches everything, so Zero-Miss, with only a tiny fraction needing your review. In seconds, you're confident that nothins is missed. No more exhaustive searches but gains in speed and trust. I think that's the future we need to build."

While he sees that with the current hype surrounding AI very high expectations toward the technology exist and its use is still frequently criticized, Reyes nevertheless has a positive outlook toward the implementation of AI tools in medical fields: "So, the good news, in my opinion, is that doctors in radiation oncology, they don't have this fear of 'AI will replace me,' those days are gone. And they do see AI as an empowering type of tool. People have expectations and they do want to see AI contributing effectively to their workflow. They don't have that mindset of AI will replace me, or AI can do everything, which is also good."

For the A-BEACON project, his wish list is clear. "It would be great that, first of all, we see the technology woring. So, all those ideas in the current actually work," Reyes elaborates. "Then, of course, I want to see that PhD student in our lab having a very nice PhD period and having a, what I call, the PhD journey. Something successful."

Primarily, however, his hopes are directed at what comes after the four-year project phase. "I hope the colleagues, they do want more, they do get engaged," the researcher muses. "I hope this is sort of the beginning, they want more, and they start, for example, reaching out to their biomedical engineering colleagues locally. And they start building up and they start even outperforming our system. That's what you want to create. Like a seed that then grows and it's bigger than what you expected. That would be, for me, the wish list."